The goal of this project is to demonstrate that incorporating biological indices into the optimization of intensity-modulation can result in a significant improvement in 3D conformal treatment plans compared to corresponding plans which are optimized based purely on dose and simple dose-volume criteria. Intensity modulation provides a greatly increased control over dose distributions. Such control can be maximally exploited to achieve the desired results using sophisticated optimization techniques. However, it is important that the objectives of optimization be specified in a clinically relevant manner in order that the results be clinically optimum. Dose-based, dose-volume based and biology-based intensity modulation optimization techniques will be applied to groups of prostate, brain and lung 3D conformal treatment plans. These plans will be compared among themselves as well as with the best non-intensity- modulated treatment plans designed by a planner. The results will be used to evaluate the relative strengths and weaknesses of various techniques and to explore the reasons for the differences between them. The intensity-modulation optimization methodology, a rudimentary version of which has been implemented in the preliminary work, will be refined. The impact of the models (serial, parallel or phenomenological) chosen to compute biological indices and their parameters on the proficiency and efficiency of the optimization process will be examined. For each class of radiotherapy problems, the parameters of the models will be adjusted to produce results that are consistent with the judgment of physicians and planners as well as with published data. Methods will be developed to accurately account for the non-local nature of radiation dose deposition in the design of optimum intensity distributions. The importance of accurate accounting of scatter on the optimality of intensity-modulated plans will be investigated. For each site and stage, the optimum number of beams for intensity-modulated treatments will be determined. The value of including non-coplanar beams will be explored. Using alpha-beta models, the potential radiobiological consequences of the various fractionation schemes involving intensity- modulated treatments will be compared. Mathematical aspects of intensity-modulation optimization methods, including multiple minima and methods to accelerate the optimization process, will be investigated. It is expected that the outcome of the proposed investigations will permit the development of intensity-modulated treatment plans in which dose to the tumor can be escalated significantly for the same or lower probability of normal tissue damage, potentially improving local control and hence, survival.